Interpolating Causal Mechanisms - The Paradox of Knowing More


Causal knowledge is not static, it is constantly modified based on new evidence. The present set of seven experiments explores one important case of causal belief revision that has been neglected in research so far - causal interpolations. A simple prototypic case of an interpolation is a situation in which we initially have knowledge about a positive covariation between two variables but later become interested in the mechanism linking these two variables. Our key finding is that interpolation tends to be misrepresented, which leads to the paradox of knowing more - The more people know about a mechanism, the less probable they tend to find the effect given the cause (i.e., weakening effect). Indeed, in all our experiments we found that, despite identical learning data about two variables, the probability linking the two variables was judged higher when follow-up research showed that the two variables were assumed to be directly causally linked (i.e., C causes E) than when subjects were instructed that the causal relation is in fact mediated by a variable representing a component of the mechanism (M) (i.e., C causes M causes E). Our explanation of the weakening effect is that people tend to confuse discoveries of pre-existing but unknown mechanisms with situations in which new variables are being added to a previously simpler causal model, which violates causal stability assumptions in natural kind domains. The experiments test several implications of this hypothesis.

In (submitted)
Simon Stephan
Research Scientist in the field of Cognitive Science at the

My research interest is computational cognitive science. I’m particularly active in the field of causal learning and causal reasoning.